Key Takeaways
- Managers of Polar Capital Technology Trust have cut back on Mag 7 holdings and instead are focusing on other stocks in the “slipstream” of AI spending.
- The trust is favoring AI infrastructure stocks in memory and power generation.
- The investment trust is up nearly 50% in share price terms this year.
Artificial intelligence investing has become increasingly concentrated, but the managers of Polar Capital Technology Trust PCT think this overlooks the companies supplying the infrastructure behind the boom.
Ben Rogoff and Alastair Unwin describe themselves as “AI maximalists,” and Nvidia NVDA is the trust’s largest holding, accounting for 10.1% of assets at the end of August. But the Gold-rated trust has reduced its exposure to the Magnificent Seven to around 30% of assets, compared with more than 50% of its benchmark, the Dow Jones Global Technology index.
Instead, they’re favoring companies benefiting from the technology giants’ capital expenditure, and finding opportunities among the companies supplying the chips, memory, networking equipment, and power needed to build AI infrastructure.
“We’re much happier to be in their slipstream,” Rogoff says. “We’re investing in companies helping to build the AI infrastructure, but not bearing the brunt of the cost.”
The £7.5 billion trust, launched in 1996 and managed by Polar Capital since 2001, invests in a diversified portfolio of listed technology and related companies. It has a Morningstar Medalist Rating of Gold.
In pound sterling, Polar Capital Technology Trust has posted a total return of more than 50% in the year to date, above the 37.78% return of the Morningstar Global Technology Index. Over three years, the trust has posted an annualized return of 44.64%, above the index return of 33.32%. And over five years, the investment trust has gained 22.75% on an annualized basis, beating the index by just over two percentage points. Polar Capital Technology Trust has also beaten 90% of its peers in the sector equity technology Morningstar category for the year-to-date, three-year, and five-year time periods.
Polar Capital typically favors companies with strong balance sheets, high margins, and defensible market positions, operating in areas where the underlying technology is moving into widespread adoption, says Michael Born, a manager research analyst at Morningstar. Born says the managers’ experience of several technology and market cycles leaves them well placed to navigate a rapidly changing sector. Their approach combines identifying disruptive technologies early with avoiding speculative businesses whose products are yet to generate meaningful revenue, he adds.
James Carthew, head of investment companies at research firm Quoted Data, says Polar’s recent performance is not simply down to its exposure to a buoyant investment theme.
“What I like is that Polar Capital Technology has done much more than just ride the AI wave, and the portfolio changes it has made have added value,” he says.
The Risk and Rewards in Owning the Magnificent Seven
The trust has held Apple AAPL continuously since 2003 and Alphabet GOOG since its IPO, and has also owned Meta Platforms META for most of its listed history.
The managers remained relatively positive on these businesses when other active investors questioned whether companies of such scale could continue growing, but they say generative AI has changed the balance between risk and reward.
The hyperscalers such as Nvidia and Alphabet are increasingly picking up the bill for AI infrastructure, while AI itself could disrupt some of the profit pools that made them dominant.
“There are ways for each of the hyperscalers to win from AI, but there are also some scenarios where they might lose, or do less well, in an AI-first world,” Rogoff explains.
Alphabet, for example, owns one of the leading AI businesses, but AI-generated answers and autonomous agents could challenge its conventional search advertising model.
“There are risks and rewards to every one of those Mag Seven stocks in a way that didn’t exist 18 months ago,” Rogoff says.
The managers see opportunities in areas previously regarded as commoditized, cyclical, or difficult places to make money. Memory is one example: The amount required in AI data centers has produced what Unwin describes as an “enormous positive demand shock,” improving volumes, pricing, and margins for suppliers.
The hunt for infrastructure bottlenecks has also taken the trust beyond conventional technology companies. Power availability is now a constraint on the development of new data centers, they say, leading Polar to invest in businesses including Siemens Energy ENR and GE Vernova GEV.
Polar Capital Technology Trust—Top Five Equity Holdings
Software Stocks and AI Disruption Risk
The other side of the repositioning has been a substantial reduction in exposure to application software stocks.
Polar previously owned companies including Adobe ADBE, Intuit INTU, and Salesforce CRM, but the managers are concerned AI could weaken some of the barriers to entry created by their large and complex codebases.
“Do you have a defensible business in a world where AI capabilities continue to improve?” Rogoff says. “That [question] has helped us hugely reduce our exposure to the software world.”
The team assesses every investment through an “AI lens,” considering whether a company enables, benefits from, or adopts the technology—or whether more capable models could undermine its business.
“We’re very sensitive to AI bear cases on our stocks, even if they seem almost fanciful at a certain point,” Unwin says. “We take them seriously, and that has kept us out of trouble.”
Is AI a Bubble?
The scale of investment pouring into AI has prompted comparisons with the dot-com boom.
Rogoff and Unwin accept that infrastructure booms can eventually produce excess capacity. But they argue that point has not yet been reached because demand for AI computing continues to outstrip supply.
“While genuine demand is outstripping supply, you can’t be in a bubble,” Unwin says. “The bubble is when you build for a world where they will come—and they don’t.”
The managers estimate annual AI-related revenue has already reached between USD 175 billion and USD 200 billion less than four years into the current cycle.
“What industry has ever got to USD 175 billion in revenue inside four years?” Rogoff says. “We are seeing the most rapid adoption of a new technology in the history of meaningful technologies.”
